One of the challenges in the Game Analytics field is to determine the type of information that can be obtained from a specific MMOG, as well as which data mining technique to use or develop depending on the peculiarities of its database. In this context, the object of study of this research is Final Fantasy XIV, a very popular MMOG that provides a limited amount of open data and also has been little researched in the literature. Therefore, this work studies the various clustering techniques as a game data mining tool. Different clustering methods are compared in order to find out the best results in the context of this game, which presents a narrow range of data for analysis. The following seven clustering algorithms were used: The k-means (partitional clustering), WARD (hierarchical), DBSCAN (density-based), spectral-based, BANG (grid-based), SOM (model-based), and Fuzzy C-means (Fuzzy Clustering). Regarding the identified player profiles by the clustering process, the results suggested the presence of five different categories: Beginner, Casual, Dedicated, Hardcore and Intermittent, characterized according to their behavior within the game. These results may contribute to a better understanding of the Final Fantasy XIV player groups and provide a basis for future work, as well as provide a case study on clustering techniques applied over a limited set of game data.